Re: [glob2-devel] Minimap algorithm

[Leo Wandersleb]

Hi Steph

As you said before, this is an open project and nobody can deside on who should 
do what. If it is Bradleys wish to improve the minimap, he should go for it. If 
he fails, he/we should revert to the old style minimap.

Sidenote: I doubt it is possible to improve the minimap significantly as i 
didn't miss a single thing about it (except for those hunger/guard/... overlays 
not going into it). I spent some thoughts on it and came to the conclusion that 
it is not trivial to both represent all enemies that maybe approach the base in 
their quantity and keep each single one notably visible. If Bradley simply 
interpolated the 512x512map-neighbors of a point in 100x100  he obviously 
misses some(most) spots.
I guess one would have to prerender all terrain features and overlay ressources 
then buildings then units/blinking markers. I guess that's what the algo was 
before.
To improve speed, one could render the terrain only once per game (as long as 
we don't allow to modify it), ressources and buildings only once every 20 
frames and units/events for every frame.

In any case please don't complain about arrogance or things like that. Keep an 
eye on the end result, point to the bug tracker if you think you can motivate 
someone to spend his tim on more urgent tasks. But please don't insult people 
as long as they don't do harm to the project (insist in merging bad code, ...). 
Your opinion on what should be merged and what reverted still counts a lot, so 
use this power wisely.

Greetings, Leo Wandersleb